CN116433312A

CN116433312A - Material pushing rule generation method and material pushing rule generation device

Info

Publication number: CN116433312A
Application number: CN202111670731.3A
Authority: CN
Inventors: 朱子晗; 白杨
Original assignee: 4Paradigm Beijing Technology Co Ltd
Current assignee: 4Paradigm Beijing Technology Co Ltd
Priority date: 2021-12-31
Filing date: 2021-12-31
Publication date: 2023-07-14

Abstract

The disclosure relates to a method and a device for generating a material pushing rule, wherein the method comprises the following steps: receiving a material pushing rule setting instruction; responding to a material pushing rule setting instruction, and displaying a setting interface of a target material pushing rule; receiving rule parameter information set in a setting interface by a user, wherein the rule parameter information comprises a client grouping identifier and/or a material grouping identifier; and generating a client grouping indicated by the identification of the client grouping and/or a target material pushing rule of the material grouping indicated by the identification of the material grouping according to the set rule parameter information. Therefore, a visual setting interface of the material pushing rules can be provided, and a user can push different types of materials for different types of people only by setting rule parameter information in the setting interface according to actual needs. The realization process for setting the material pushing rule is simple, the labor cost and the time cost can be saved, and the method is convenient and quick.

Description

Material pushing rule generation method and material pushing rule generation device

Technical Field

The present disclosure relates to the field of computer technologies, and in particular, to a method and an apparatus for generating a material pushing rule.

Background

For various intelligent marketing platforms, materials can be pushed to clients. For example, the e-commerce platform may push clothing-like materials, food-like materials, or daily-use-like materials to customers; the information platform can push information-like materials to clients, etc. If the material pushing strategy is formulated, larger benefits can be brought to the marketing platform, so that how to set the material pushing rules is important.

In the related art, if different types of materials are to be pushed for different types of people, a special technical development capability is required, however, an operator of the intelligent marketing platform may not have such a capability, and at this time, a special developer is required to write a script and modify a code for different situations. Moreover, due to the diversity of crowd types and the diversity of material types, a large number of combinations between crowd types and material types can occur. At this time, a developer is required to develop a corresponding code for each combination of a large number of combinations, so that the code development workload of pushing different types of materials for different types of people is large, and large labor cost and time cost are consumed.

Disclosure of Invention

The disclosure provides a method and a device for generating a material pushing rule, so as to at least solve the problem that in the related art, if different types of materials are to be pushed to different types of people, developers are required to develop corresponding codes for each combination in a large number of combinations, and larger labor cost and time cost are consumed.

According to a first aspect of embodiments of the present disclosure, there is provided a method for generating a material pushing rule, applied to a server, the method including: receiving a material pushing rule setting instruction; responding to the material pushing rule setting instruction, and displaying a setting interface of a target material pushing rule; receiving rule parameter information set by a user in the setting interface, wherein the rule parameter information comprises a client grouping identifier and/or a material grouping identifier; and generating a customer cluster indicated by the identification of the customer cluster and/or the target material pushing rule of the material cluster indicated by the identification of the material cluster according to the set rule parameter information.

Optionally, before the step of receiving the material pushing rule setting instruction, the method further includes: receiving a material grouping setting instruction; and setting a plurality of material groups in response to the material group setting instruction, wherein each material group in the plurality of material groups corresponds to an identification of the material group and a material group rule, and the material group rule is a related rule, an interval rule or a subordinate rule.

Optionally, the correlation rule is used for searching materials, of which the corresponding description information is matched with the relevant information of the client, from a plurality of description information corresponding to the materials stored in the server, as materials contained in the material group carrying the correlation rule.

Optionally, the interval rule is used for searching a first operation result of the corresponding description information from a plurality of description information corresponding to a plurality of materials stored in the server, and materials with second operation results of the related information of the client meeting a preset size relationship are used as materials contained in the material group carrying the interval rule; the first operation result is a four-rule operation result between the description information and a first constant, the second operation result is a four-rule operation result between the relevant information of the client and a second constant, and the preset size relation comprises at least one of greater than, greater than or equal to, less than and less than or equal to.

Optionally, the subordinate rule is used for searching materials with inclusion relation between the corresponding description information and the relevant information of the client from a plurality of description information corresponding to a plurality of materials stored in the server, and the materials are used as materials contained in the material group loaded with the subordinate rule.

Optionally, the generating, according to the set rule parameter information, the target material pushing rule for the customer cluster indicated by the identification of the customer cluster and/or the material cluster indicated by the identification of the material cluster includes: and generating a pushing prohibition rule for the material group so as to prohibit pushing the materials contained in the material group to the terminals of the clients contained in the client group.

Optionally, the generating, according to the set rule parameter information, the target material pushing rule for the customer cluster indicated by the identification of the customer cluster and/or the material cluster indicated by the identification of the material cluster includes: and generating a top-setting display rule for the material group so that the materials contained in the material group are top-set and displayed on the terminals of the clients contained in the client group.

Optionally, the generating, according to the set rule parameter information, the target material pushing rule for the customer cluster indicated by the identification of the customer cluster and/or the material cluster indicated by the identification of the material cluster includes: and generating a diversity rule for the material group so as to limit the occurrence frequency of the materials contained in the material group on the terminals of the clients contained in the client group to be not more than a preset frequency.

Optionally, the generating, according to the set rule parameter information, the target material pushing rule for the customer cluster indicated by the identification of the customer cluster and/or the material cluster indicated by the identification of the material cluster includes: and generating a push rule for the material group so as to force the materials contained in the material group to be pushed to the terminals of the clients contained in the client group.

Optionally, the generating, according to the set rule parameter information, the target material pushing rule for the customer cluster indicated by the identification of the customer cluster and/or the material cluster indicated by the identification of the material cluster includes: and generating a weight adjustment rule for the material group so as to adjust the display weight of the material contained in the material group on the terminal of the customer contained in the customer group, wherein the larger the display weight of the material is, the more preferentially the material is displayed.

Optionally, the rule parameter information further includes a plurality of model weights corresponding to a plurality of estimated models accessed by the server; generating the client group indicated by the identification of the client group and/or the target material pushing rule of the material group indicated by the identification of the material group according to the set rule parameter information, including: and generating a weighted summation rule for each material contained in the material group by utilizing a plurality of pre-estimated values corresponding to the plurality of pre-estimated models and the plurality of model weights, wherein the larger the weighted summation result corresponding to the material is, the more preferentially the material is pushed.

Optionally, the rule parameter information further includes a plurality of model weights corresponding to a plurality of estimated models accessed by the server; generating the client group indicated by the identification of the client group and/or the target material pushing rule of the material group indicated by the identification of the material group according to the set rule parameter information, including: and generating a weighted product rule for each material contained in the material group by utilizing a plurality of pre-estimated values corresponding to the plurality of pre-estimated models and the plurality of model weights, wherein the larger the weighted product result corresponding to the material is, the more preferentially the material is pushed.

Optionally, the plurality of estimation models are a customer stay time estimation model, a click probability estimation model, a collection shopping probability estimation model and a conversion probability estimation model.

According to a second aspect of the embodiments of the present disclosure, there is provided a device for generating a material pushing rule, including: the receiving module is configured to receive a material pushing rule setting instruction; the display module is configured to respond to the material pushing rule setting instruction and display a setting interface of a target material pushing rule; the receiving module is configured to receive rule parameter information set by a user in the setting interface, wherein the rule parameter information comprises an identification of customer clusters and/or an identification of material clusters; and the generation module is configured to generate the client group indicated by the identification of the client group and/or the target material pushing rule of the material group indicated by the identification of the material group according to the set rule parameter information.

Optionally, the generating device further comprises a setting module; the receiving module is configured to receive a material grouping setting instruction; the setting module is configured to respond to the material grouping setting instruction to set a plurality of material groupings, wherein each material grouping in the plurality of material groupings corresponds to an identification of the material grouping and a material grouping rule, and the material grouping rule is a correlation rule, an interval rule or a subordinate rule.

Optionally, the generating module is configured to: and generating a pushing prohibition rule for the material group so as to prohibit pushing the materials contained in the material group to the terminals of the clients contained in the client group.

Optionally, the generating module is configured to: and generating a top-setting display rule for the material group so that the materials contained in the material group are top-set and displayed on the terminals of the clients contained in the client group.

Optionally, the generating module is configured to: and generating a diversity rule for the material group so as to limit the occurrence frequency of the materials contained in the material group on the terminals of the clients contained in the client group to be not more than a preset frequency.

Optionally, the generating module is configured to: and generating a push rule for the material group so as to force the materials contained in the material group to be pushed to the terminals of the clients contained in the client group.

Optionally, the generating module is configured to: and generating a weight adjustment rule for the material group so as to adjust the display weight of the material contained in the material group on the terminal of the customer contained in the customer group, wherein the larger the display weight of the material is, the more preferentially the material is displayed.

Optionally, the rule parameter information further includes a plurality of model weights corresponding to a plurality of estimated models accessed by the server; the generation module is configured to: and generating a weighted summation rule for each material contained in the material group by utilizing a plurality of pre-estimated values corresponding to the plurality of pre-estimated models and the plurality of model weights, wherein the larger the weighted summation result corresponding to the material is, the more preferentially the material is pushed.

Optionally, the rule parameter information further includes a plurality of model weights corresponding to a plurality of estimated models accessed by the server; the generation module is configured to: and generating a weighted product rule for each material contained in the material group by utilizing a plurality of pre-estimated values corresponding to the plurality of pre-estimated models and the plurality of model weights, wherein the larger the weighted product result corresponding to the material is, the more preferentially the material is pushed.

According to a third aspect of embodiments of the present disclosure, there is provided an electronic device, comprising: a processor; a memory for storing the processor-executable instructions; the processor is configured to execute the instructions to implement a method for generating a material pushing rule according to the disclosure.

According to a fourth aspect of embodiments of the present disclosure, there is provided a computer readable storage medium, which when executed by a processor of an electronic device, causes the electronic device to perform a method of generating a stock pushing rule according to the present disclosure.

The technical scheme provided by the embodiment of the disclosure at least brings the following beneficial effects:

the visual setting interface of the material pushing rules can be provided, and a user can push different types of materials for different types of people only by setting rule parameter information in the setting interface according to actual needs. The client grouping can be used as a main function of client grouping management and calculation, the material grouping can be used as a main function of object grouping management and calculation, and the client grouping and the material grouping are called in the operation scene material pushing rule configuration, so that complicated person and object selection in the material pushing rule configuration is realized. The method has the advantages that a developer does not need to develop corresponding codes respectively aiming at a large number of combinations between crowd types and material types, zero-code and low-cost manual intervention of recommending the whole process can be realized, instant, personalized and flexible material pushing rules are realized for service call, and the implementation process of setting the material pushing rules is simple, convenient and quick.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the disclosure and together with the description, serve to explain the principles of the disclosure and do not constitute an undue limitation on the disclosure.

FIG. 1 is a flow chart illustrating a method of generating a material pushing rule according to an exemplary embodiment of the present disclosure;

FIG. 2 is a schematic diagram illustrating a setup interface for a material push rule for a material selection phase according to an exemplary embodiment of the present disclosure;

FIG. 3 is a schematic diagram illustrating a setup interface for a material pushing rule for a material rearrangement phase according to an exemplary embodiment of the present disclosure;

FIG. 4 is a schematic diagram illustrating relevant information of one or more predictive models in accordance with an exemplary embodiment of the present disclosure;

FIG. 5 is a schematic diagram illustrating an interface for adding a predictive model in accordance with an exemplary embodiment of the present disclosure;

fig. 6 is a block diagram illustrating a generation apparatus of a material pushing rule according to an exemplary embodiment of the present disclosure;

fig. 7 is a block diagram illustrating an electronic device according to an exemplary embodiment of the present disclosure.

Detailed Description

In order to enable those skilled in the art to better understand the technical solutions of the present disclosure, the technical solutions of the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.

It should be noted that the terms "first," "second," and the like in the description and claims of the present disclosure and in the foregoing figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the disclosure described herein may be capable of operation in sequences other than those illustrated or described herein. The embodiments described in the examples below are not representative of all embodiments consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present disclosure as detailed in the accompanying claims.

It should be noted that, in this disclosure, "at least one of the items" refers to a case where three types of juxtaposition including "any one of the items", "a combination of any of the items", "an entirety of the items" are included. For example, "including at least one of a and B" includes three cases side by side as follows: (1) comprises A; (2) comprising B; (3) includes A and B. For example, "at least one of the first and second steps is executed", that is, three cases are juxtaposed as follows: (1) performing step one; (2) executing the second step; (3) executing the first step and the second step.

The intelligent marketing platform can be used for three intelligent engines: intelligent recommendation, intelligent search and intelligent pushing are realized, intelligent operation of a plurality of scenes is realized rapidly, and core service indexes such as guest price, user retention, repurchase rate, total amount of deals (Gross Merchandise Volume, GMV) and the like are improved. If an enterprise wants to access the recommended service, the method can be divided into four steps: uploading material data, creating recommendation service, acquiring recommendation results and uploading user behavior data.

(1) Uploading material data

Materials refer to recommended materials used by recommended services, such as articles, merchandise, pictures, and the like. Material data is the basis for the recommended service, so it is first necessary to transfer the material data into the material warehouse of the system. The material data is used for calculating material portraits, model training and feature service to perform feature calculation.

(2) Creating recommendation services

When creating the recommended service, selecting a material library for service call; a recommendation scene template for use by the service is selected, e.g., a popular recommendation, a personalized recommendation, an associated recommendation, etc. After the creation is completed, the system can automatically connect all links in the recommendation process in series, and set corresponding strategies and models in a recall stage, a coarse arrangement stage, a fine arrangement stage and a rearrangement stage according to the selected recommendation algorithm.

(3) Acquiring recommended results

When a user requests a recommendation service, the recommendation service firstly distributes flow according to the parameter transmission in the request, then carries out a recall stage, a filtering and screening stage, a coarse arrangement stage, a fine arrangement stage and a rearrangement stage, and finally gives a recommendation result to the user.

The whole recommendation flow can be as follows:

a. the client refreshes the screen once at the terminal, which corresponds to initiating a Request (Request) to the server.

b. And (3) entering a Recall (Recall) stage, recalling materials from a material library by using different algorithms through multiple recalls, fusing (Merge) the recalled materials, performing filtering (Filter) such as duplicate removal and the like. Wherein, recall calculation is applied to material portraits and user portraits.

c. The personalized model ordering is participated, and coarse Rank (Prerank), fine Rank (Rank) and rearrangement (Rerank) are adopted in the ordering. The recalled and filtered materials can be respectively sequenced for a plurality of times, so that the material scale is gradually reduced, the model dimension is increased, and the personalized fineness is improved. The coarse ranking calculation is to roughly sort and cut off recall results according to a small amount of characteristic data; the fine ranking calculation is to accurately rank the feature data provided by the feature service by using a complex artificial intelligence (Artificial Intelligence, AI) model service; the rearrangement calculation is integrated with various business rules, and the ranking list obtained by the fine rearrangement calculation is adjusted to obtain a final recommendation return result.

d. And returning the recommended final result from the big-to-small arrangement according to the probability of clicking and other actions generated by the user.

(4) Uploading user behavior data

After the user obtains the recommendation result, corresponding actions such as exposure, clicking, collection, praise and the like can occur, and the action data of the user can be uploaded to the server.

After the four steps are completed, the product can provide end-to-end recommendation service, and a customer enterprise can acquire the personalized recommendation result returned for the terminal user each time only by calling a recommendation service interface.

It should be noted that, in the related art, if different types of materials are to be pushed for different types of people, it is required to have a professional technical development capability, however, for an operator of the intelligent marketing platform, it may not have such a capability, and at this time, a special developer is required to write a script and modify a code for different situations. Moreover, due to the diversity of crowd types and the diversity of material types, a large number of combinations between crowd types and material types can occur. At this time, a developer is required to develop a corresponding code for each combination of a large number of combinations, so that the code development workload of pushing different types of materials for different types of people is large, and large labor cost and time cost are consumed.

In order to solve the above problems in the related art, the method for generating the material pushing rule provided by the present disclosure may provide a visual setting interface of the material pushing rule, and a user may push different types of materials for different types of people only by setting rule parameter information in the setting interface according to actual needs. The client grouping can be used as a main function of client grouping management and calculation, the material grouping can be used as a main function of object grouping management and calculation, and the client grouping and the material grouping are called in the operation scene material pushing rule configuration, so that complicated person and object selection in the material pushing rule configuration is realized. The method has the advantages that a developer does not need to develop corresponding codes respectively aiming at a large number of combinations between crowd types and material types, zero-code and low-cost manual intervention of recommending the whole process can be realized, instant, personalized and flexible material pushing rules are realized for service call, and the implementation process of setting the material pushing rules is simple, convenient and quick.

Fig. 1 is a flowchart illustrating a method of generating a material pushing rule according to an exemplary embodiment of the present disclosure, which is applied to a server.

Referring to fig. 1, in step 101, a material push rule setting instruction may be received. It should be noted that the material pushing rule may be divided into three stages: the material selecting stage, the material sorting stage and the material rearranging stage respectively correspond to a recall stage and a filtering stage of materials, a model sorting stage (a coarse arranging stage and a fine arranging stage) and a rearranging stage.

In the concrete floor of the recommendation/push scenario, the following traffic scenario requirements are often encountered:

in the maternal and infant commodity recommendation, after the month age is formulated by a user, milk powder with the same month age as the user is required to be screened as a recommendation result; in the maternal and infant commodity recommendation, after a user purchases milk powder, the user is required to screen expert questions and answers related to the titles of the purchased commodities, wherein "purchase" refers to "purchase vehicle; in financial recommendation, users have risk levels and are required to screen products with the risk level not higher than the current user; in stock information recommendation, a user often holds stock in a stock warehouse and requests to screen information related to the stock titles held in the stock warehouse by the user; in the recommendation of the meal, a user holds a meal card ticket and is required to screen out the meal which is the same as the coupon to stimulate consumption.

The above scenes show the variability of commodity screening in the actual scene, and the screening itself represents the constraint condition of the business level, especially in the financial field, with stronger compliance requirements. In the field of retail goods, the screening conditions include signals of business to facilitate the operation. Meanwhile, when screening is performed, logic judgment screening is often performed depending on the context of the user (including environmental information and a dynamic representation of the user).

In addition, in the concrete floor of recommended scenes, another type of business scene requirement is often encountered:

in retail recommendation, after the model completes personalized sorting according to the purchase conversion rate, rearrangement is required according to the price of the commodity from high to low; in the information recommendation, after the geographic information is uploaded by a user, the food information is recommended from near to far according to the requirement; in retail recommendation, it is desirable to reorder the results (often with lower prices followed by higher prices) sorted by purchase conversion to a price per purchase conversion to promote GMV; in the information class recommendation, the model ordering can be interfered, so that the model ordering can be used as the model ordering basis according to the click rate and the browsing market.

The above scenes show variability of commodity ordering in actual scenes, and the result screened out according to the fixed model can not completely meet the service requirement. Therefore, different indexes are required to be used as sequencing basis of a sequencing model according to different scenes, and even the weight mixing of multiple indexes is required to be used for promoting the core service indexes of the current scene.

According to an exemplary embodiment of the present disclosure, a material group setting instruction may be received, and then, in response to the material group setting instruction, a plurality of material groups may be set. Each material grouping in the plurality of material groupings can correspond to a material grouping identifier and a material grouping rule, and the material grouping rule can be a related rule, an interval rule or a subordinate rule.

It should be noted that a plurality of client groups may be provided. Wherein the clients included in each of the plurality of client clusters are of the same type. For example, a grouping rule for female client grouping may be set as: if the sex of the user is female, after receiving a material pushing request sent by a certain client, if the sex of the client is determined to be female, the client can be brought into the female client grouping; alternatively, a grouping rule for grouping non-married clients may be set as: if the marital status of the user is not married after receiving the material pushing request sent by a certain client, the client can be incorporated into the 'not married client grouping'.

According to an exemplary embodiment of the present disclosure, the correlation rule is used to find, from a plurality of description information corresponding to a plurality of materials stored in the server, materials whose corresponding description information matches with the relevant information of the client, as materials contained in the material group carrying the correlation rule. For example, the correlation rule may be configured to:

"Material (1) field is related to user (2) field"

Wherein "related" is defined as: the query is obscured. Specific logic of "correlation" is: after the value of a certain field of the material is segmented, fuzzy matching is carried out on the segmented value of the certain field of the user and the content of the segmented value of the certain field of the user, and if matching is successful, the material is screened; otherwise, the screening is not performed.

Here:

(1) the option may be to take a field from a physical attribute (containing a portrait attribute), which only supports an array class field, e.g., a shopping cart list field, or a string class field. The field support of the material: basic properties of the material and portrait properties of the material. Wherein, the basic attribute of material contains: title (title), content (content), category1-5, tag (tag); the portrait attributes of the material include: material classification probability (item Category), material title keyword (item title kw), material content keyword (item ContentKw).

(2) A field of the user may be selected that only supports an array class field, such as a shopping cart list field or a string class field. The fields of the user may support: user interest (interest), subscription (subscale), material title of last 50 clicks (title 50), material title of last 50 clicks (title 1 m), material title of last 50 clicks (title Kw 50), material title of last 50 clicks (title Kw1 m), material content of last 50 clicks (content Kw 50), material content of last 50 clicks (content Kw1 m), material tag of last 50 clicks (tag 50), material tag of last 1 month click (tag 1 m), material category of last 50 clicks (category 50), material category of last 50 clicks (category 1 m), material category of last 50 clicks (category 150), secondary material category of last 50 clicks (category 250), tertiary material category of last 50 clicks (category 350), tertiary material category of last 21 clicks (category of last 21 months clicks), tertiary material category of last 21 clicks (category of last 21 months, etc.

For example, the following dynamic grouping rules may be set in a physical property scenario:

"the last three days of purchase of a user is related to the label of the material" represents that the user wishes to screen out financial information whose label is related to the last three days of purchase of stock in the user portrait.

According to an exemplary embodiment of the present disclosure, an interval rule is used to search for, from a plurality of description information corresponding to a plurality of materials stored in a server, a first operation result of the corresponding description information, and a material whose second operation result of related information of a client satisfies a preset size relationship, as a material included in a material group carrying the interval rule.

The first operation result may be a four-rule operation result between the description information and the first constant; the second operation result can be a four-rule operation result between the related information of the client and the second constant; the preset size relationship may include at least one of greater than, equal to or less than, and less than or equal to.

For example, the interval rule may be configured to:

"Material (1) field (2) plus/minus/multiply/divide (3) constant value (4) greater than/equal to/less than user (5) field"

Here:

(1) the field may be selected from the object properties (including the image properties) and only the value type field is supported;

(2) add/subtract/multiply/divide may be selected;

(3) a constant value can be selected, the range can be 1-100, and the precision can be 2-bit decimal;

(4) five comparison operators as described above may be selected;

(5) a certain field of the user attribute (including the portrait attribute) may be selected and only the numeric class field is supported by the field.

For example, in a point mall good redemption scenario, an operator configures two dynamic grouping rules as:

"(integral of material multiplied by 0.5) is less than or equal to (integral of user)".

"equal to or greater than (integral of the material multiplied by 1) (integral of the user multiplied by 1)".

Then the customer would like to screen the material for which the redemption points are between 1-2 times the user points.

According to an exemplary embodiment of the present disclosure, a subordinate rule is used to find, from a plurality of description information corresponding to a plurality of materials stored in a server, materials in which a corresponding description information has an inclusion relationship with related information of a client, as materials included in a material group bearing the subordinate rule.

For example, the slave rule may be configured to:

"Material (1) field (2) included/contained in user (3) field"

Here:

(1) the field can be selected from the object attribute (including the image attribute), and the field can support all the array type fields or the character string type fields;

(2) can be selected to be included in, and subject to subordinate relation selection;

(3) a certain field of the user attribute (portrait attribute) may be selected and only a group class field or a string class field is supported by the field.

For example, "the user's baby month age is included in the available month ages of the material" means that the available baby month ages of the mother and infant goods required to be screened need to include the user's baby month ages.

In this way, operators such as 'containing', 'correlation', 'equal' and the like aiming at characters can be designed when setting a material grouping rule of dynamic grouping; operators such as ' greater than ', ' greater than or equal to ', ' less than or equal to ', addition, subtraction, multiplication and division ' aiming at numerical values can be designed, operation logic can be enriched, multiple calculation modes can be flexibly supported, and the problem that multiple fields are difficult to match and operate is solved.

In step 102, in response to the material pushing rule setting instruction, a setting interface of the target material pushing rule may be displayed.

In step 103, rule parameter information set by a user in a setting interface may be received, where the rule parameter information may include an identification of a customer cluster and/or an identification of a material cluster.

In step 104, a target material pushing rule for the customer cluster indicated by the identification of the customer cluster and/or the material cluster indicated by the identification of the material cluster may be generated according to the set rule parameter information.

It should be noted that the material selection stage may have a screening rule and a candidate rule for forward selecting materials, and a limiting rule and a deduplication rule for filtering materials. The material pushing rules of the material selection stage act on the recall stage and the filtering stage, and are used for screening out a material corpus meeting the service requirements, wherein the material corpus participates in subsequent sorting.

The material pushing rules for positively selecting materials according to business logic are as follows:

screening rules:

in order to accept the requirement of grouping according to the appointed materials as a full-quantity material set, group screening can be set, and the bold marked part can be configured:

"recall (1) only the material that appears in a material cluster for all users" is used to accept the screening of a particular cluster using dynamic clusters.

Wherein, the liquid crystal display device comprises a liquid crystal display device,

(1) a certain material grouping can be selected, and the material grouping supports dynamic grouping and static grouping. It should be noted that "static grouping" refers to that material grouping is performed by using a single attribute or multiple attributes of materials in advance, and then updating is performed for different material groupings periodically. For example, material groupings may be updated on a daily basis.

For example, "recall only the material that appears in the clothing material cluster for all users" represents that the operator wishes to screen clothing material as a complete set of recalls. The number of each rule scheme in the filtering rule scene can be set according to practical situations, for example, up to five rule schemes can be supported.

Candidate rule:

in order to accept the requirement that recall can be newly added based on the business rule, a candidate rule can be set, namely, certain materials can be clustered to be used as one recall and other recalls in parallel, the fused materials are used as a candidate material set, and the bold marked part can be configured:

"all users-oriented newly added recall (1)" grouping something "

Here:

(1) one or more material clusters may be selected, and the material clusters may support dynamic clusters and static clusters.

For example:

the customer configuration ' recall user real-time interests, mother and infant, cosmetic material grouping for all users ' new increases ', represents hope to add materials in user real-time interests material grouping, mother and infant material grouping, cosmetic material grouping into recall complete set after screening.

The candidate material grouping rule corresponding to the user real-time interest material grouping can be that the description information of the materials is related to the user interests, and the user real-time interest material grouping can be obtained by searching the materials marked with the mountain climbing label from the materials on the premise that the user interests are determined to be mountain climbing after the material pushing request sent by the terminal of the user is received; the candidate material grouping rule corresponding to the mother and infant material grouping can be that the description information of the materials is related to the mother and infant, and the materials marked with the mother and infant labels can be searched from a plurality of materials to form the mother and infant material grouping; the candidate material grouping rule corresponding to the 'cosmetic material grouping' can be that the description information of the materials is related to the cosmetic, and the materials with the 'cosmetic' labels can be searched out from a plurality of materials to form the cosmetic material grouping.

It should be noted that, the candidate rule has no limit on the number of the candidate rules, and in the material pushing rule of the forward selected material, the effective priority of the screening rule is greater than the effective priority of the candidate rule, that is, the recalled material of the candidate rule must satisfy the screening rule. If the same material hits a candidate rule but does not hit a screening rule, then the material is not recommended. When a plurality of screening rules exist, the materials hit by the screening rules are intersected, namely when some materials do not meet all the screening rules, the materials are not pushed.

Therefore, the candidate rule realizes the effect of adding, namely materials which are required to be pushed to the terminal of the client can be added into the recall corpus through the candidate rule; the filtering rule realizes the filtering effect, namely the materials recalled by the candidate rule can be filtered through the filtering rule, and only the filtered materials meeting the filtering rule are used as the materials which are finally pushed to the terminal of the client.

According to an exemplary embodiment of the present disclosure, the material pushing rule of the material (filtered material) is reversely selected according to the business logic as a push prohibition rule. A push prohibition rule may be generated for the material cluster to prohibit pushing of the material contained in the material cluster to the terminal of the client contained in the client cluster. Further, the push prohibition rule may be classified into a deduplication rule and a restriction rule.

And (3) a duplication removal rule:

to accommodate the need for "no recommendation for a period of time that a user's exposed/clicked merchandise is required" a deduplication rule may be set, and the bold marked portion may be configured to:

"Material deduplication is performed for (1) a certain user grouping/all users using (2) actions such as (3) clicking/exposing occur within xx days, effective time: (4) xx year xx month xx day xx hour xx score-xx year xx month xx day xx hour xx score/permanent effectiveness'

Here:

(1) the existing user group (supporting the ID of the group or the name of the group for inquiring) under a certain current tenant can be selected, or all users can be selected to be targeted;

(2) a time window range of 1-31 days can be selected, and the unit of the day can be adopted;

(3) the duplicate removal actions that can be selected are: click/buy/expose/click (click) etc., where "click" refers to a bad evaluation of the material.

(4) The validation time range may be selected to be: all times are permanently effective, or the earliest time can be the current time, or the initial default time window in the customization can be in the range of the current time to 24 hours later (year-month-day-minute-year-month-day-minute).

One per scenario may be supported per scheme for the deduplication rule. For example, the set deduplication rule may be: in the recommendation of the E-commerce, the commodity purchased by the user is de-duplicated; alternatively, "in information recommendation, deduplication of user-exposed information"

Restriction rules:

in order to meet the requirement of 'requiring underage and not recommending smoke and wine materials', a limit rule can be set to realize blackening:

"presentation of (1) a user grouping/all user restrictions (2) a material grouping

Time of validation: (3) xx year xx month xx day xx hour xx score-xx year xx month xx day xx hour xx score/permanent effectiveness'

Here:

(2) one or more material groups (supporting group IDs or group names for inquiry) can be selected, and a single material (supporting material IDs for inquiry) can be selected;

(3) the validation time range may be selected to be: all times are permanently effective, or the time is custom (year-month-day-minute-year-month-day-minute, earliest time can be the current time, and the initial default time window in the custom can be in the range of the current time to 24 hours later.

The effective priority of the limiting rules is highest, once the limiting rules are configured, all other material pushing rules can be covered, when one client hits a plurality of limiting rules at the same time, the limiting rules are effective at the same time, and the limiting rules have no number limitation.

It should be noted that, the effective priority of the material pushing rule of the reverse selection material is higher than the effective priority of the material pushing rule of the forward selection material: when the same material is hit by the material pushing rule of the reverse selection material and the material pushing rule of the forward selection material, the material pushing rule of the reverse selection material is effective. In the material pushing rule of the reverse selection material, the effective priority of the limiting rule is higher than the effective priority of the duplicate removal rule.

Referring to fig. 2, fig. 2 is a schematic diagram illustrating a setting interface of a material push rule of a material selection stage according to an exemplary embodiment of the present disclosure. In fig. 2, the setting interfaces of the four material pushing rules of "candidate rule", "filtering rule", "limiting rule" and "duplication removing rule" are shown. Under the "candidate rule", 4 examples are listed, namely "1 and facing the user", and the user can be recommended to click on the materials of the related candidate clusters in three days, wherein the candidate ratio is 50%; 2, the user can be recommended to browse the materials which are clustered by the related candidates in three days, and the candidate ratio is 20%; 3, the user can be recommended to purchase the materials which are clustered by the related candidates in three days, and the candidate ratio is 10%; and 4, recommending materials recalled by the user vector similar strategy by the user facing the user, wherein the candidate ratio is 20%. Wherein, the "candidate duty ratio" refers to the proportion of the materials recalled by the corresponding candidate rule in the recall total set.

Under the "screening rule", 1 example is listed: 1, recommending only commodity grouping materials which are matched with the age of a user in a grading way for the user; under the "restriction rule", 1 example is listed: 1, recommending materials with expired effective time for all users; under the "deduplication rule", 1 example is listed: "1, for all users, does not recommend purchased materials.

According to an exemplary embodiment of the present disclosure, the material rearrangement stage acts on the material sorting stage in service logic, and both reorders the model sorting results, so as to display the sorting results according to service requirements. The material rearrangement stage involves a set-top display rule, a push rule, a diversity rule and a weight adjustment rule.

Set-top display rules:

the top-setting display rule can be generated for material grouping, so that the materials contained in the material grouping are displayed on the terminals of the clients contained in the client grouping, and the bold marked part can be configured:

common overhead display:

1) "in the time frame of (1) xx, facing (2) xx crowd, top (3) xx Material in the 4 th xx request (5) xx position"

For example: "at 2021-4-10:00-2021-4-20:23, facing all users, overhead two related information content, at the first location of the first request of the users. ", representing that it is desired to display two parties of related content information at the first location of each request during two parties.

Permanent set-top display:

2) "in the time frame of (1) xx, facing (2) xx crowd, permanently set top (3) xx Material in the position of (4) xx"

For example: at 2021-4-10 00:00-2021-4-20:59, facing all users, the permanent set-top "content_id=00243" is in the first position of the "refresh" request, representing that it is desirable to permanently set-top the article of content_id=00243 during the two parties.

It should be noted that the validation priority of the set-top display rule is only inferior to the validation priority of the limit rule in the material selection stage. If a certain material is hit by the limiting rule, the top-mounted display rule is not executed; if a material is not hit by the constraint rule, the set-top display rule may be executed. When the user hits a plurality of top-setting display rules at the same time and generates position conflict, the top-setting display rules which are newly configured can be validated; or, hit a plurality of and set up and reveal the rule with users at the same time, and produce the position conflict, can point out "there are other materials set up in the current position already". The set top display rules may support only 2 bars per scheme per scene.

According to an exemplary embodiment of the present disclosure, a push-must rule may be generated for a material cluster to force a material contained in the material cluster to be pushed to a terminal of a customer contained in a customer cluster, and the bold-marked portion may be configured to:

"force to reveal (2) a material grouping/single material for (1) a user grouping/all users, effective time: (3) xx year xx month xx day xx hour xx score-xx year xx month xx day xx hour xx score/permanent effectiveness'

Here:

(2) one or more material groupings may be selected (support for grouping ID or grouping name for querying), or a single material may be selected (support for querying by material ID);

The effective priority of the push rule is lower than that of the set-top display rule, and the push rule can be executed after the push rule is satisfied by recommending content every time a user brushes. Meanwhile, the effective priority of the rule to be pushed is lower than that of all the material pushing rules in the material selection stage, namely, the effective priority of the rule to be pushed is lower than that of the candidate rule, the screening rule, the limit rule and the duplicate removal rule in the material selection stage. When a user requests that the materials are not in the final screening range screened by the material pushing rule in the material selection stage at one time, but hit by the pushing rule, the pushing rule is not executed. When the user hits a plurality of rule to be pushed at the same time and generates position conflict, the latest rule to be pushed can be used as the standard, and the rule to be pushed has no limit on the number of the rules.

According to an exemplary embodiment of the present disclosure, in order to suppress the presentation of a single product material at the same recommendation site, for example, to accommodate the requirement that "no more than 5 lipstick materials are required to be recommended for beauty lovers", a diversity rule may be generated for material grouping to limit the number of occurrences of the material contained in the material grouping on the terminals of the clients contained in the client grouping to be no more than a preset number of times, and the bold marked portion may be configured to:

"display of (1) a certain user grouping/all users limit (2) a certain material grouping no more than (3) a certain number of times, effective time: (4) xx year xx month xx day xx hour xx score-xx year xx month xx day xx hour xx score/permanent effectiveness'

Here:

(1) existing user grouping (supporting grouping ID or grouping name for inquiring) under a certain current tenant can be selected, or all users can be selected to be targeted;

(2) one or more material clusters (ID of supporting cluster or name of cluster for inquiry) can be selected;

(3) a positive integer may be selected, for example, in the range of 1 to 10;

(4) the validation time range may be selected to be: all times are permanently effective, or the time is custom (year-month-day-minute-year-month-day-minute, earliest time can be the current time, and the initial default time window in the custom can be in the range of the current time to 24 hours later.

The validation priority of the diversity rule is lower than the validation priority of the must-push rule. When multiple diversity rules exist simultaneously, the minimum number is taken for the materials hit the multiple diversity rules simultaneously. For example: the material collection comprises materials A, B and C, and the diversity rule corresponding to the material collection is not more than 1; the other material set comprises a material B, a material C and a material D, and the diversity rule corresponding to the material set is not more than 2. At this time, the materials contained in the intersection of the two material sets are the material B and the material C, and then the diversity rule corresponding to the intersection takes 1 smaller piece. The diversity rule is number-free.

According to an exemplary embodiment of the present disclosure, in order to accommodate the requirement that "certain materials need to be weighted for certain user groups," a weighting rule may be generated for material groupings to adjust the presentation weight of the materials contained in the material groupings on the terminals of the clients contained in the client groupings. Wherein, the larger the display weight of the material, the more preferentially the material is displayed. The bold marked portions may be configured to:

"for (1) a certain user grouping, adjusting the material display weight in (2) a certain material grouping to (3)1-5) a certain value, and the effective time: (4) xx year xx month xx day xx hour xx score-xx year xx month xx day xx hour xx score/permanent effectiveness'

Wherein:

(2) one or more material clusters (ID supporting the cluster or name of the cluster) may be selected for querying.

It should be noted that the weight adjustment rule only supports one scene configuration.

Referring to fig. 3, fig. 3 is a schematic diagram illustrating a setting interface of a material pushing rule of a material rearrangement stage according to an exemplary embodiment of the present disclosure. In fig. 3, setting interfaces of "set-top display rule", "push rule", "diversity rule" and "weight adjustment rule" in the material rearrangement stage are shown. Among them, the "set-top display rule" lists an example: "1," group top display of two news materials group for all users, time range: the year 2020, the month 07, the month 01, the day 8 and the year 2020, the month 07, the day 20 and the day 0 are respectively; an example is listed below "must push rule": "1," face the presentation of the forced XX new product material grouping of beauty fan user grouping, time frame: the year 2020, the month 07, the month 01, the day 8 and the year 2020, the month 07, the day 20 and the day 0 are respectively; an example is listed below for the "diversity rule: "1," military fan user-oriented grouping limit military news material grouping presentation no more than 2 times, time range: permanently validated "; an example of a "weighting rule" is listed below: "1," XX user-oriented grouping adjusts the showing weight of the materials in the news material grouping to 2, and the time range is: permanently effective).

According to an exemplary embodiment of the present disclosure, the rule parameter information may further include a plurality of model weights corresponding to a plurality of estimated models accessed by the server. And generating a weighted summation rule for each material contained in the material group by utilizing a plurality of pre-estimated values and a plurality of model weights corresponding to the plurality of pre-estimated models, wherein the larger the weighted summation result corresponding to the material is, the more preferentially the material is pushed.

For example, the weighted summation rule may be:

(Pre-estimated model 1 weight) x P ₁ ++ (pre-estimated model 2 weight) ×P ₂ ++ (predictive model 3 weight). Times.P ₂ +…

Wherein P is ₁ The scoring value of the material by the pre-estimating model 1 is meant; p (P) ₂ The scoring value of the material by the pre-estimation model 2 is meant; p (P) ₃ Refers to scoring of the material by the predictive model 3, and so on.

According to an exemplary embodiment of the present disclosure, the rule parameter information may further include a plurality of model weights corresponding to a plurality of pre-estimated models accessed by the server, and the weighted product rule may be generated for each material included in the material group by using a plurality of pre-estimated values corresponding to the plurality of pre-estimated models and the plurality of model weights. The larger the weighted product result corresponding to the material is, the more preferentially the material is pushed.

For example, the weighted product rule is as follows:

According to an exemplary embodiment of the present disclosure, the plurality of estimation models may be a customer stay time estimation model, a click probability estimation model, a collection shopping probability estimation model, a conversion probability estimation model, and the like. Referring to fig. 4, fig. 4 is a schematic diagram illustrating related information of one or more predictive models according to an exemplary embodiment of the present disclosure. In FIG. 4, the name, type, and model address of each of the customer stay time prediction model, the click probability prediction model, the collection purchase probability prediction model, and the conversion probability prediction model are shown.

It should be noted that, the material sorting stage of the present disclosure may support zero code addition of the prediction model, and only needs to configure the address of the prediction model. Referring to fig. 5, fig. 5 is a schematic diagram illustrating an interface for adding a predictive model according to an exemplary embodiment of the present disclosure. In fig. 5, an input box for adding the name of the predictive model is shown, and two types of predictive model options are provided for the operator to select, respectively "external predictive model service" and "predictive model service". Therefore, the generation method of the material pushing rule can realize zero code addition pre-estimated model, and the implementation process is simple, convenient and quick. According to the method, the device and the system, the pre-estimated index values of different pre-estimated models can be used as sequencing basis according to the needs of different scenes, further, different pre-estimated index values of different pre-estimated models can be weighted and calculated according to different weights to sequence materials, better sequencing effect can be obtained, and further, larger benefits can be brought to an intelligent marketing platform.

Further, the method for generating the material pushing rule of the present disclosure may further set a bottom protection rule, where the bottom protection rule functions as follows: if the materials are screened by using the various material pushing rules, the material which can be pushed to the terminal of the client is not obtained, and then the bottom protection rule can be executed. For example, the bottom policy may be set to: "recommended latest product for all users". Therefore, by setting the bottom protection rule, when materials which can be pushed to the terminal of the client are not screened out, the feedback to the client can be ensured, and bad use experience caused by directly returning information of failed searching the materials is avoided.

Fig. 6 is a block diagram illustrating a generation apparatus of a material pushing rule according to an exemplary embodiment of the present disclosure.

Referring to fig. 6, the apparatus 600 may include a receiving module 601, a display module 602, and a generating module 603.

A receiving module 601 configured to receive a material pushing rule setting instruction;

the display module 602 is configured to respond to the material pushing rule setting instruction and display a setting interface of a target material pushing rule;

the receiving module 601 is configured to receive rule parameter information set by a user in the setting interface, where the rule parameter information includes an identifier of a customer cluster and/or an identifier of a material cluster;

The generating module 603 is configured to generate, according to the set rule parameter information, the target material pushing rule for the customer cluster indicated by the identification of the customer cluster and/or for the material cluster indicated by the identification of the material cluster.

According to an exemplary embodiment of the present disclosure, the generating apparatus 600 may further include a setting module;

the receiving module 601 is configured to receive a material grouping setting instruction;

the setting module is configured to respond to the material grouping setting instruction to set a plurality of material groupings, wherein each material grouping in the plurality of material groupings corresponds to an identification of the material grouping and a material grouping rule, and the material grouping rule is a correlation rule, an interval rule or a subordinate rule.

According to an exemplary embodiment of the present disclosure, the correlation rule is used to find, from a plurality of description information corresponding to a plurality of materials stored in the server, a material whose corresponding description information matches with the relevant information of the client, as a material contained in a material group carrying the correlation rule.

According to an exemplary embodiment of the present disclosure, the interval rule is configured to search, from a plurality of description information corresponding to a plurality of materials stored in the server, a first operation result of the corresponding description information, and a material, in which a second operation result of related information of a client satisfies a preset size relationship, as a material included in a material group carrying the interval rule;

The first operation result is a four-rule operation result between the description information and a first constant, the second operation result is a four-rule operation result between the relevant information of the client and a second constant, and the preset size relation comprises at least one of greater than, greater than or equal to, less than and less than or equal to.

According to an exemplary embodiment of the present disclosure, the subordinate rule is configured to search, from a plurality of description information corresponding to a plurality of materials stored in the server, materials having an inclusion relationship between the corresponding description information and relevant information of a client, as materials included in a material group carrying the subordinate rule.

According to an exemplary embodiment of the present disclosure, the generating module 603 is configured to:

and generating a pushing prohibition rule for the material group so as to prohibit pushing the materials contained in the material group to the terminals of the clients contained in the client group.

and generating a top-setting display rule for the material group so that the materials contained in the material group are top-set and displayed on the terminals of the clients contained in the client group.

and generating a diversity rule for the material group so as to limit the occurrence frequency of the materials contained in the material group on the terminals of the clients contained in the client group to be not more than a preset frequency.

and generating a push rule for the material group so as to force the materials contained in the material group to be pushed to the terminals of the clients contained in the client group.

and generating a weight adjustment rule for the material group so as to adjust the display weight of the material contained in the material group on the terminal of the customer contained in the customer group, wherein the larger the display weight of the material is, the more preferentially the material is displayed.

According to an exemplary embodiment of the present disclosure, the rule parameter information further includes a plurality of model weights corresponding to a plurality of estimated models accessed by the server;

the generation module 603 is configured to:

and generating a weighted summation rule for each material contained in the material group by utilizing a plurality of pre-estimated values corresponding to the plurality of pre-estimated models and the plurality of model weights, wherein the larger the weighted summation result corresponding to the material is, the more preferentially the material is pushed.

the generation module 603 is configured to:

and generating a weighted product rule for each material contained in the material group by utilizing a plurality of pre-estimated values corresponding to the plurality of pre-estimated models and the plurality of model weights, wherein the larger the weighted product result corresponding to the material is, the more preferentially the material is pushed.

According to an exemplary embodiment of the present disclosure, the plurality of estimation models are a customer stay time estimation model, a click probability estimation model, a collection purchase probability estimation model, and a conversion probability estimation model.

Fig. 7 is a block diagram illustrating an electronic device 700 according to an exemplary embodiment of the present disclosure.

Referring to fig. 7, an electronic device 700 includes at least one memory 701 and at least one processor 702, the at least one memory 701 having instructions stored therein that, when executed by the at least one processor 702, perform a method of generating a material pushing rule according to an exemplary embodiment of the present disclosure.

By way of example, the electronic device 700 may be a PC computer, tablet device, personal digital assistant, smart phone, or other device capable of executing the instructions described above. Here, the electronic device 700 is not necessarily a single electronic device, but may be any apparatus or a collection of circuits capable of executing the above-described instructions (or instruction set) individually or in combination. The electronic device 700 may also be part of an integrated control system or system manager, or may be configured as a portable electronic device that interfaces with either locally or remotely (e.g., via wireless transmission).

In electronic device 700, processor 702 may include a Central Processing Unit (CPU), a Graphics Processor (GPU), a programmable logic device, a special purpose processor system, a microcontroller, or a microprocessor. By way of example, and not limitation, processors may also include analog processors, digital processors, microprocessors, multi-core processors, processor arrays, network processors, and the like.

The processor 702 may execute instructions or code stored in the memory 701, wherein the memory 701 may also store data. The instructions and data may also be transmitted and received over a network via a network interface device, which may employ any known transmission protocol.

The memory 701 may be integrated with the processor 702, for example, RAM or flash memory disposed within an integrated circuit microprocessor or the like. In addition, the memory 701 may include a separate device, such as an external disk drive, a storage array, or any other storage device usable by a database system. The memory 701 and the processor 702 may be operatively coupled or may communicate with each other, for example, through an I/O port, a network connection, etc., such that the processor 702 is able to read files stored in the memory.

In addition, the electronic device 700 may also include a video display (such as a liquid crystal display) and a user interaction interface (such as a keyboard, mouse, touch input device, etc.). All components of the electronic device 700 may be connected to each other via a bus and/or a network.

According to an exemplary embodiment of the present disclosure, a computer-readable storage medium may also be provided, which when executed by a processor of an electronic device, enables the electronic device to perform the above-described method of generating a material pushing rule. Examples of the computer readable storage medium herein include: read-only memory (ROM), random-access programmable read-only memory (PROM), electrically erasable programmable read-only memory (EEPROM), random-access memory (RAM), dynamic random-access memory (DRAM), static random-access memory (SRAM), flash memory, nonvolatile memory, CD-ROM, CD-R, CD + R, CD-RW, CD+RW, DVD-ROM, DVD-R, DVD + R, DVD-RW, DVD+RW, DVD-RAM, BD-ROM, BD-R, BD-R LTH, BD-RE, blu-ray or optical disk storage, hard Disk Drives (HDD), solid State Disks (SSD), card memory (such as multimedia cards, secure Digital (SD) cards or ultra-fast digital (XD) cards), magnetic tape, floppy disks, magneto-optical data storage, hard disks, solid state disks, and any other means configured to store computer programs and any associated data, data files and data structures in a non-transitory manner and to provide the computer programs and any associated data, data files and data structures to a processor or computer to enable the processor or computer to execute the programs. The computer programs in the computer readable storage media described above can be run in an environment deployed in a computer device, such as a client, host, proxy device, server, etc., and further, in one example, the computer programs and any associated data, data files, and data structures are distributed across networked computer systems such that the computer programs and any associated data, data files, and data structures are stored, accessed, and executed in a distributed fashion by one or more processors or computers.

According to the method and the device for generating the material pushing rules, a visual material pushing rule setting interface can be provided, and a user can push different types of materials for different types of people only by setting rule parameter information in the setting interface according to actual needs. The client grouping can be used as a main function of client grouping management and calculation, the material grouping can be used as a main function of object grouping management and calculation, and the client grouping and the material grouping are called in the operation scene material pushing rule configuration, so that complicated person and object selection in the material pushing rule configuration is realized. The method has the advantages that a developer does not need to develop corresponding codes respectively aiming at a large number of combinations between crowd types and material types, zero-code and low-cost manual intervention of recommending the whole process can be realized, instant, personalized and flexible material pushing rules are realized for service call, and the implementation process of setting the material pushing rules is simple, convenient and quick. Further, when setting a material grouping rule of dynamic grouping, operators such as "contain", "related", "equal" and the like aiming at characters can be designed; operators such as ' greater than ', ' greater than or equal to ', ' less than or equal to ', addition, subtraction, multiplication and division ' aiming at numerical values can be designed, operation logic can be enriched, multiple calculation modes can be flexibly supported, and the problem that multiple fields are difficult to match and operate is solved. Furthermore, the method for generating the material pushing rule can realize zero code addition estimation model, and is simple in realization process, convenient and quick. According to the method, the device and the system, the pre-estimated index values of different pre-estimated models can be used as sequencing basis according to the needs of different scenes, further, different pre-estimated index values of different pre-estimated models can be weighted and calculated according to different weights to sequence materials, better sequencing effect can be obtained, and further, larger benefits can be brought to an intelligent marketing platform. Further, by setting the bottom protection rule, when materials which can be pushed to the terminal of the client are not screened out, the feedback to the client can be ensured, and bad use experience caused by directly returning information of failed searching the materials is avoided.

Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This disclosure is intended to cover any adaptations, uses, or adaptations of the disclosure following the general principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

It is to be understood that the present disclosure is not limited to the precise arrangements and instrumentalities shown in the drawings, and that various modifications and changes may be effected without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. The method for generating the material pushing rule is applied to a server and is characterized by comprising the following steps:

receiving a material pushing rule setting instruction;

responding to the material pushing rule setting instruction, and displaying a setting interface of a target material pushing rule;

receiving rule parameter information set by a user in the setting interface, wherein the rule parameter information comprises a client grouping identifier and/or a material grouping identifier;

And generating a customer cluster indicated by the identification of the customer cluster and/or the target material pushing rule of the material cluster indicated by the identification of the material cluster according to the set rule parameter information.

2. The method of generating a material push rule according to claim 1, wherein prior to the step of receiving a material push rule setting instruction, the method further comprises:

receiving a material grouping setting instruction;

and setting a plurality of material groups in response to the material group setting instruction, wherein each material group in the plurality of material groups corresponds to an identification of the material group and a material group rule, and the material group rule is a related rule, an interval rule or a subordinate rule.

3. The method for generating a material pushing rule according to claim 2, wherein the correlation rule is used for searching, from a plurality of description information corresponding to a plurality of materials stored in the server, for a material whose corresponding description information matches with the relevant information of the client, as a material included in the material group carrying the correlation rule.

4. The method for generating a material pushing rule according to claim 2, wherein the interval rule is used for searching a first operation result of the corresponding description information from a plurality of description information corresponding to a plurality of materials stored in the server, and a material with a second operation result of the related information of the client meeting a preset size relationship is used as a material contained in a material group carrying the interval rule;

5. The method for generating a rule for pushing materials according to claim 2, wherein the subordinate rule is used for searching materials having a containing relationship between the corresponding description information and the relevant information of the client from a plurality of description information corresponding to a plurality of materials stored in the server as materials contained in the material group carrying the subordinate rule.

6. The method according to claim 1, wherein the generating the target material pushing rule for the customer group indicated by the identification of the customer group and/or the material group indicated by the identification of the material group according to the set rule parameter information includes:

7. The method according to claim 1, wherein the generating the target material pushing rule for the customer group indicated by the identification of the customer group and/or the material group indicated by the identification of the material group according to the set rule parameter information includes:

8. The device for generating the material pushing rule is characterized by comprising the following components:

the receiving module is configured to receive a material pushing rule setting instruction;

the display module is configured to respond to the material pushing rule setting instruction and display a setting interface of a target material pushing rule;

the receiving module is configured to receive rule parameter information set by a user in the setting interface, wherein the rule parameter information comprises an identification of customer clusters and/or an identification of material clusters;

and the generation module is configured to generate the client group indicated by the identification of the client group and/or the target material pushing rule of the material group indicated by the identification of the material group according to the set rule parameter information.

9. An electronic device, comprising:

a processor;

a memory for storing the processor-executable instructions;

wherein the processor is configured to execute the instructions to implement the method of generating a material push rule according to any one of claims 1 to 7.

10. A computer readable storage medium, characterized in that instructions in the computer readable storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the method of generating a material push rule according to any one of claims 1 to 7.